Cell lines from cancer patients who undergo radiation therapy have been correlated to clinical outcomes and to in vitro radiation responses. That some cells of similar histologies, particularly from patients whose treatments have failed to locally control their tumors, show in vitro radiation survival parameters of a resistant type, while others are of a sensitive type, suggests a genetic basis for radiation sensitivity. The overall objective of this project is to identify genes that may be associated with the radiation resistance of human tumors. Using the techniques of protein analysis (2-D gel electrophoresis and microsequencing) differential screening/subtraction cDNA library analysis, and vector mediated gene transfer, a multiple approach strategy is planned in the first AIM to identify and isolate genes associated with the radiation-resistant phenotype. We will test the hypothesis that differential gene expressions in cells are responsible for the radiation sensitive or the radiation resistant phenotypes of tumor cells. In the second AIM of this project, two strategies to identify genes associated with resistance will be used to modify the radiation responses of sensitive tumor cells using vector mediated gene transfer. The first will use an episomally maintained vector strategy containing cDNA from radiation resistant cells and the second will use a cosmid strategy for gene transfer using larger fragments of DNA from resistant cells. Transfections will be performed into radiation sensitive human tumor cells and phenotypic selection will be performed to permit subsequent rescue of the DNA fragments. The identification of differentially expressed genes in radiation resistant cells will provide insight into the mechanisms of tumor radiation resistance and may be a future basis for tissue based predictive tests.